An Unmanned Aerial Vehicle Scheduling Method and System for Air-Ground Collaboration

Abstract

The present invention relates to the technical field of unmanned aerial vehicles. More specifically, the present invention relates to a method and system for unmanned aerial vehicle scheduling for air-ground cooperation, including: collecting the power consumption and wind speed of the unmanned aerial vehicle at the same moment; dividing the wind speed collected at each moment along the time line into a stable wind speed group and an abnormal wind speed group based on the volatility of the wind speed, establishing a mapping relationship between the stable wind speed group and the power consumption, and calculating the unmanned aerial vehicle scheduling safety factor at the current moment, and performing scheduling based on the magnitude of the unmanned aerial vehicle scheduling safety factor at the current moment. The present invention uses a clustering algorithm to screen the real-time collected wind speed, screens out abnormal wind speed data and eliminates it, reduces the interference of the abnormal wind speed data on the mapping relationship between the power consumption and the wind speed, and improves the accuracy of calculating the unmanned aerial vehicle scheduling safety factor subsequently.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicles. More specifically, the present invention relates to a method and system for unmanned aerial vehicle scheduling for air-ground cooperation. Background Art

[0002] With the development of unmanned aerial vehicle technology, it plays an important role in air-ground cooperation tasks. An air-ground cooperative unmanned aerial vehicle refers to the use of information interaction and cooperative control between a ground terminal (such as a dispatching center, an unmanned aerial vehicle ground station, an intelligent vehicle, etc.) and an unmanned aerial vehicle to achieve functions such as remote monitoring, intelligent scheduling, energy replenishment, and task optimization of the unmanned aerial vehicle. The aerial part (unmanned aerial vehicle) is responsible for task execution, such as inspection, surveying and mapping, material delivery, etc. The ground part (intelligent terminal / vehicle / control center) is responsible for calculation, monitoring, navigation assistance, data fusion, energy replenishment, etc. Moreover, during the information collection process of the unmanned aerial vehicle, the power consumption is often accelerated due to wind interference, and the actual flight time is much lower than the theoretical value. For example, when flying against the wind, the motor needs to increase power to maintain speed. At the same time, the wind causes attitude fluctuations, and the flight control system needs to correct frequently at a high frequency, further increasing power consumption. For example, the instantaneous change in wind speed will cause the instantaneous power consumption of the unmanned aerial vehicle to rise.

[0003] In the prior art, when an unmanned aerial vehicle flies in a complex wind speed environment, the power consumption of the unmanned aerial vehicle is difficult to accurately predict. And during the flight of the unmanned aerial vehicle, the influence of wind speed change on the power consumption of the unmanned aerial vehicle is significant, and abnormal wind speed data will interfere with the mapping relationship between power consumption and wind speed. Therefore, when accurately evaluating the return time and power safety of the unmanned aerial vehicle, abnormal wind speed data may cause errors in the safety evaluation of the unmanned aerial vehicle, resulting in the risk that the unmanned aerial vehicle may crash due to insufficient power. Summary of the Invention

[0004] The present invention provides a method and system for unmanned aerial vehicle scheduling for air-ground cooperation, aiming to solve the problem that in the related art, the influence of wind speed change on the power consumption of the unmanned aerial vehicle is significant, and abnormal wind speed data will interfere with the mapping relationship between power consumption and wind speed. Therefore, errors may occur when accurately evaluating the return time and power safety of the unmanned aerial vehicle.

[0005] In a first aspect, the present invention provides a method for unmanned aerial vehicle scheduling for air-ground cooperation, including: collecting the power consumption and wind speed of the unmanned aerial vehicle at the same moment; dividing the collected wind speed fluctuations at each moment along the time line into a stable wind speed group and an abnormal wind speed group, establishing a mapping relationship between the stable wind speed group and the power consumption, and calculating the unmanned aerial vehicle scheduling safety factor at the current moment, and performing scheduling based on the magnitude of the unmanned aerial vehicle scheduling safety factor at the current moment; wherein, calculating the unmanned aerial vehicle scheduling safety factor at the current moment , and the calculation formula is: ; in the formula, is the number of abnormal wind speed data within the target time period, is the number of valid data at the current moment, is the remaining battery power of the UAV at the current moment, is the time required for the UAV to return to the ground terminal, is the acquisition frequency, For the th moment within the target time period, the power consumption corresponding to the collected wind speed, where the number of valid data is related to the wind speed magnitude at the current moment and all the data within and between the two consecutive stable wind speed groups before the current moment. The target time period is the time period composed of all the moments between the current moment and its previous moments. is the number of moments within the target time period, where the number of valid data is related to the wind speed magnitude at the current moment and all the data within and between the two consecutive stable wind speed groups before the current moment. The target time period is the time period composed of all the moments between the current moment and its previous moments. By mapping the stable wind speed group and the power consumption, the stable wind speed group and the abnormal wind speed group are distinguished, the interference of the drastic wind speed fluctuation on the prediction is reduced, the accuracy of the power consumption prediction is improved, and the actual power consumption of the UAV at different wind speeds is calculated by using the historical wind speed data, avoiding the deviation caused by simply relying on the theoretical model.

[0006] Furthermore, scheduling is performed based on the magnitude of the UAV scheduling safety factor at the current moment, including: if the UAV scheduling safety factor at the current moment is less than the safety threshold, a request for return warning is sent to the ground terminal, where the empirical value of the safety threshold is 0.3.

[0007] Furthermore, calculate the number of valid data at the current moment, and the calculation formula is ; where, is the number of valid data at the current moment, is the wind speed at the current moment, is the maximum stable wind speed of the UAV, is the number of all wind speeds within and between the two consecutive stable wind speed groups before the current moment, where the maximum stable wind speed is set by the UAV's own parameters. By this method, the number of valid data can be adjusted in real time according to the wind speed magnitude at the current moment, and the number of valid data used to predict the UAV wind speed can be effectively controlled.

[0008] Further, the wind speeds collected at each moment are divided into multiple wind speed groups, including: starting from the wind speed collected at the \(i\)-th moment for clustering. Among them, the clustering method can use the K-means clustering method. For each additional moment, the response function value of the clustering is calculated once; if the response function value is less than the clustering threshold, the clustering is completed, and a clustering cluster is obtained. One clustering cluster corresponds to one wind speed group, and all the wind speeds collected between the \(i\)-th moment and the moment when the response function value is less than the clustering threshold are included in the clustering cluster. When the wind speed changes violently, the response function value drops rapidly, and the clustering ends, so that the mutant wind speeds form separate groups; when the wind speed changes smoothly, the clustering continues to ensure that the stable wind speed data can be grouped into the same group.

[0009] Further, the wind speed groups are divided into stable wind speed groups and abnormal wind speed groups, including: judging whether a wind speed group is abnormal according to the number of wind speeds in the wind speed group; if the number of wind speeds in the wind speed group is less than the number threshold, it is determined that the wind speed group is an abnormal wind speed group; if the number of wind speeds in the wind speed group is greater than or equal to the number threshold, it is determined that the wind speed group is a stable wind speed group. Distinguishing between stable wind speed groups and abnormal wind speed groups based on the number of wind speed data within the wind speed group requires no complex calculations and is suitable for real-time wind speed monitoring.

[0010] Further, the calculation method of the response function value of the wind speed at each moment is as follows: ; where is the response function value of the wind speed at the -th moment, is the sequence composed of all wind speed data between the -th moment and the -th moment, is the wind speed at the -th moment, is the exponential function with the natural constant as the base, is the mean function, STD is the standard deviation, where the standard deviation reflects the volatility of the wind speed. This method combines the mean and the standard deviation to measure the wind speed change, and the exponential function adaptively adjusts the sensitivity to improve the recognition accuracy of abnormal wind speeds, taking into account both short-term and long-term changes, reducing misjudgments, and improving the accuracy of subsequent calculations.

[0011] Further, a mapping relationship between the stable wind speed group and the power consumption is established, including: obtaining the power consumption and the normal wind speed at each moment in the stable wind speed group, and obtaining a scatter plot established by the power consumption and the normal wind speed at each moment; using the spline fitting tool in the matlab tool to fit the relationship curve between the normal wind speed and the power consumption, so as to establish the mapping relationship between the power consumption and the normal wind speed.

[0012] In a second aspect of the present invention, there is also provided a drone scheduling system for air-ground collaboration, including a processor and a memory. The memory stores a computer program, and the processor executes the computer program to implement the drone scheduling method for air-ground collaboration as described in any one of the above.

[0013] Beneficial effects:

[0014] (1) Using a clustering algorithm to screen the wind speeds collected in real time, screening out abnormal wind speed data and eliminating it, reducing the interference of abnormal wind speed data on the mapping relationship between power consumption and wind speed, and improving the accuracy of calculating the safety factor of drone scheduling subsequently.

[0015] (2) The volatility of wind speed directly affects the power consumption of drones. By establishing a wind speed-power consumption mapping relationship based on a stable wind speed group, this method can more accurately predict future power consumption, and can adaptively adjust the prediction model under different wind speed conditions, improving the accuracy of estimating the return time and power safety of drones, reducing the risk of power shortage caused by sudden wind speed changes, and helping to optimize the mission planning and scheduling strategies of drones. Description of the drawings

[0016] Figure 1 It is a flowchart showing schematically the calculation of the safety factor of drone scheduling according to an embodiment of the present invention. Detailed implementation manners

[0017] The following will describe in detail the specific implementation manners of the present invention with reference to the drawings.

[0018] As Figure 1 shown, S101: Collect the power consumption and wind speed of the drone at the same moment.

[0019] In one embodiment, during a flight mission of a drone, collect the real-time power consumption and real-time wind speed of the drone (through an ultrasonic anemometer carried by the drone). One collection moment includes one wind speed and one power consumption (the power difference between adjacent collection moments).

[0020] S102: Obtain a stable wind speed group.

[0021] In one embodiment, since when the wind speed changes significantly, whether the wind speed decreases or increases, it will cause the fuselage of the drone to fluctuate violently, resulting in an increase in the instantaneous power consumption of the drone, and this situation will interfere with the determination of the mapping relationship between real-time power consumption and real-time wind speed. Therefore, it is necessary to eliminate this abnormal data to reduce the interference of abnormal wind speed on the mapping relationship between power consumption and wind speed.

[0022] Specifically, based on the volatility of the wind speed collected at each moment along the time line, it is divided into a stable wind speed group and an abnormal wind speed group, where the abnormal wind speed group is abnormal data. It should be noted that the process of dividing the wind speed group is as follows: cluster the real-time collected wind speed, and specifically use the K-means clustering method for clustering. Specifically, start clustering with the wind speed collected at the i-th moment. For each additional moment, calculate the response function value of the clustering once. If the response function value of the added moment is less than the clustering threshold, complete this clustering to obtain a clustering cluster, and one clustering cluster corresponds to one wind speed group, where all the wind speeds collected between the i-th moment and the moment when the response function value is less than the clustering threshold (the added moment) are included in the clustering cluster. Then start a new clustering again from the next moment of the added moment according to the above method. Thus, the wind speed group can be divided based on the real-time collected wind speed. Among them, in this embodiment, the empirical value of the clustering threshold is 0.4. In other embodiments, the empirical value of the clustering threshold can be 0.5 or 0.44, etc., and can be adjusted according to the specific implementation situation.

[0023] Exemplarily, starting from the wind speed collected at the i-th moment, the response function value of the wind speed at the i + 1-th moment is calculated to be 0.5. Since the response function value at the i + 1-th moment is greater than the clustering threshold, the i + 2-th moment is added to continue clustering. The response function value of the wind speed at the i + 2-th moment is 0.6. Since the response function value at the i + 2-th moment is also greater than the clustering threshold, the i + 3-th moment is added to continue clustering. The response function value of the wind speed at the i + 3-th moment is 0.45. Since the response function value at the i + 3-th moment is also greater than the clustering threshold, the i + 4-th moment is added to continue clustering. The response function value of the wind speed at the i + 4-th moment is 0.3. At this time, the response function value at the i + 4-th moment is less than the clustering threshold, so the clustering is completed, and the wind speeds collected from the i-th moment to the i + 4-th moment are determined as one wind speed group. Then start a new clustering from the i + 5-th moment.

[0024] In one embodiment, the calculation method of the response function value of the wind speed at each moment is: ; In the formula, is the response function value of the wind speed at the -th moment, is the sequence composed of all wind speed data between the -th moment and the -th moment, is the wind speed at the -th moment, is the exponential function with the natural constant as the base, is the mean function, and STD is the standard deviation, where the standard deviation reflects the volatility of the wind speed.

[0025] Among them, is the stability of all wind speeds between the moment and the moment. The larger this value is, the more stable the wind speed change is between the moment and the moment. The more the wind speed between the moment and the moment can be classified into a clustering cluster. is the overall difference between the wind speed at the moment and all wind speeds between the moment and the moment, that is, the overall change amplitude of the wind speed at the moment relative to all wind speeds between the moment and the moment. The larger this value is, the less the moment belongs to the clustering cluster formed by all wind speeds between the moment and the moment.

[0026] As mentioned above, after dividing into wind speed groups, it is possible to determine whether the wind speed group belongs to a stable wind speed group or an abnormal wind speed group according to the number of wind speeds in the wind speed group. Specifically, if the number of wind speeds in the wind speed group is less than the number threshold, it is determined that the wind speed group is an abnormal wind speed group; if the number of wind speeds in the wind speed group is greater than or equal to the number threshold, it is determined that the wind speed group is a stable wind speed group, where the empirical value of the number threshold is 15. The reason is that: when the wind speed fluctuates, it is instantaneous rather than continuous, and each wind speed in the wind speed group corresponds to a moment. Therefore, it is possible to judge the duration of the wind speed according to the number of wind speeds in the wind speed group. When the number in the wind speed group is less than the number threshold, it means that the wind speed changes suddenly in a short time, and the wind speed group is marked as an abnormal wind speed group. On the contrary, when the number of wind speeds in the clustering cluster is greater than or equal to the number threshold, it means that the clustering cluster is a stable wind speed group.

[0027] It should be noted that: the determination method of the number threshold can be determined according to the specifications and performance of the unmanned aerial vehicle. Exemplarily: the balance recovery time of an ordinary unmanned aerial vehicle affected by sudden wind speed changes is 1.5 seconds. If the wind speed acquisition frequency is 10 hz, the data number threshold is 15. Therefore, the wind speed group greater than or equal to the number threshold is determined as a stable wind speed group.

[0028] S103: Calculate the unmanned aerial vehicle scheduling safety factor.

[0029] In one embodiment, based on the stable wind speed group, the mapping relationship between power consumption and normal wind speed is obtained using Matlab, that is, the power consumption of the drone under different wind speeds can be obtained. Specifically, the power consumption and normal wind speed at each moment in the stable wind speed group are obtained, and then a scatter plot established by the power consumption and normal wind speed at each moment is obtained. Using the spline fitting tool in the Matlab tool, the relationship curve between the normal wind speed and power consumption is fitted, thereby establishing the mapping relationship between power consumption and normal wind speed. Then, referring to the occurred wind speed, the decrease amplitude of the drone's power consumption in a future period of time is predicted to determine whether the drone can safely return during the return time. Specifically, the time for the drone to return to the ground vehicle is obtained, and the power consumption of the drone during this time is predicted, thereby calculating the drone scheduling safety factor.

[0030] In one embodiment, calculate the drone scheduling safety factor at the current moment , and the calculation formula is: ; In the formula, is the number of abnormal wind speed data in the target time period, is the number of valid data at the current moment, is the remaining power of the drone at the current moment, is the time required for the drone to return to the ground terminal, is the acquisition frequency, is the power consumption corresponding to the wind speed collected at the th moment in the target time period, where the number of valid data is related to the magnitude of the wind speed at the current moment and all the data within and between the two consecutive stable wind speed groups before the current moment. The target time period is the time period composed of all the moments between the current moment and the preceding moments, is the number of moments in the target time period, and the target time period is the time period composed of all the moments between the current moment and the preceding moments.

[0031] Among them, is the average power consumption of the drone obtained within the range of the above-mentioned number of valid data, which is used as the real-time power consumption during the time when the drone returns to the ground vehicle in the future. At the same time, it is compared with the remaining power to obtain the power danger coefficient of the drone. The larger this value is, the more the remaining power is relatively, the smaller the return risk of the drone is, and the higher the drone scheduling safety factor is. It is used to limit the number of abnormal wind speed data within the range of valid data. When the number of abnormal wind speed data is relatively large, the UAV dispatch safety factor should be actively reduced, because the more abnormal wind speed data, the more unstable the wind speed is, and the average consumption rate of the remaining power will not be greater than the average power consumption of the UAV within the range of valid data. Therefore, the larger the value, the fewer the number of abnormal wind speed data within the range of valid data, indicating that the power consumption predicted by the UAV is more accurate.

[0032] In one embodiment, the number of valid data at the current moment is related to the wind speed at the current moment and the number of all data within and between two consecutive stable wind speed groups before the current moment. The method for determining the number of valid data at the current moment needs to satisfy the following relationship: ;in, is the number of valid data at the current moment, is the wind speed at the current moment, is the maximum stable wind speed of the drone, is the number of all wind speeds within and between two consecutive stable wind speed groups before the current moment, where the maximum stable wind speed is set by the drone's own parameters, and the wind speed value range at the current moment should be less than the maximum stable wind speed, because when the wind speed at the current moment is equal to or greater than the maximum stable wind speed of the drone, it means that the drone can no longer maintain balance and may fall, and the drone should be alarmed at this time. The number of valid data That is to say, the number of all wind speeds within and between two consecutive stable wind speed groups before the current moment will be affected by the real-time wind speed. The closer the wind speed at the current moment is to the maximum stable wind speed of the drone, the more likely the balance of the drone will be threatened. Therefore, for safety reasons, the number of valid data used to predict the wind speed of the drone should be reduced.

[0033] In one embodiment, the number of all wind speeds within and between two consecutive stable wind speed groups before the current moment is obtained. , for example: the wind speed groups divided before the current moment are stable wind speed group Z1, abnormal wind speed group Z2, stable wind speed group Z3 and stable wind speed group Z4 in order from near to far in time. The two consecutive stable wind speed groups before the current moment are stable wind speed group Z1 and stable wind speed group Z3, so the number The number includes the number of wind speeds in the stable wind speed group Z1 and the stable wind speed group Z3, and also includes the number of wind speeds in the abnormal wind speed group Z2.

[0034] S104: Perform scheduling based on the drone scheduling safety factor.

[0035] In one embodiment, if the UAV scheduling safety factor at the current moment is less than the safety threshold, a request for return warning is sent to the ground terminal. Among them, the empirical value of the safety threshold is 0.3. In other embodiments, the empirical value of the safety threshold can be 0.33 or 0.45, etc.

[0036] The present invention also provides a UAV scheduling system for air-ground cooperation. The system includes a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a UAV scheduling method according to the first aspect of the present invention is implemented.

[0037] The system also includes other components well-known to those skilled in the art, such as a communication bus and a communication interface. Their settings and functions are known in the art, so they will not be elaborated here.

[0038] In the present invention, the aforementioned memory can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, device, or component. For example, the computer-readable storage medium can be any suitable magnetic storage medium or magneto-optical storage medium, such as resistive random access memory (RRAM), dynamic random access memory (DRAM), static random access memory (SRAM), enhanced dynamic random access memory (EDRAM), high-bandwidth memory (HBM), hybrid memory cube (HMC), etc., or any other medium that can be used to store the required information and can be accessed by an application, module, or both. Any such computer storage medium can be part of the device or accessible or connectable to the device. Any application or module described in the present invention can be implemented using computer-readable / executable instructions stored or otherwise maintained on such a computer-readable medium.

[0039] The above-described embodiments merely represent several implementation manners of the present invention. Their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the patent application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention.

Claims

1. A method for dispatching unmanned aerial vehicles for air-ground collaboration, characterized in that: include: Collect the power consumption and wind speed of the drone at the same time; Based on the volatility of wind speed collected at each moment along the timeline, it is divided into a stable wind speed group and an abnormal wind speed group, a mapping relationship between the stable wind speed group and power consumption is established, and the drone dispatch safety factor at the current moment is calculated, and the dispatch is performed based on the size of the drone dispatch safety factor at the current moment; Among them, the safety factor of drone dispatch at the current moment is calculated , the calculation formula is: ； In the formula, is the number of abnormal wind speed data in the target time period, is the number of valid data at the current moment, is the remaining power of the drone at the current moment, The time required for the drone to return to the ground terminal. is the acquisition frequency, The target time period The power consumption corresponding to the wind speed is collected at each moment, wherein the number of valid data is related to the wind speed at the current moment and the number of all data within and between two consecutive stable wind speed groups before the current moment, and the target time period is the current moment and the previous moment. The time period consists of all the moments between is the number of time points in the target time period.

2. The method for dispatching unmanned aerial vehicles for air-ground collaboration according to claim 1, characterized in that: The dispatch is performed based on the current UAV dispatch safety factor, including: If the UAV dispatch safety factor at the current moment is less than the safety threshold, a return warning request is sent to the ground terminal.

3. The method for dispatching unmanned aerial vehicles for air-ground collaboration according to claim 1, characterized in that: Calculate the number of valid data at the current moment. The calculation formula is: ； in, is the number of valid data at the current moment, is the wind speed at the current moment, is the maximum stable wind speed of the drone, It is the number of all wind speeds within and between two consecutive stable wind speed groups before the current moment, where the maximum stable wind speed is set by the drone's own parameters.

4. The method for dispatching unmanned aerial vehicles for air-ground collaboration according to claim 1, characterized in that: The wind speed collected at each moment is divided into multiple wind speed groups, including: Clustering starts with the wind speed collected at the i-th moment, and the clustering response function value is calculated once each time a moment is added; If the response function value is less than the clustering threshold, the clustering is completed to obtain a cluster, and one cluster corresponds to one wind speed group, wherein the cluster contains all wind speeds collected between the i-th moment and the moment when the response function value is less than the clustering threshold.

5. The method for dispatching unmanned aerial vehicles for air-ground collaboration according to claim 4, characterized in that: The wind speed group is divided into a stable wind speed group and an abnormal wind speed group, including: Judging whether the wind speed group is abnormal according to the number of wind speeds in the wind speed group; If the number of wind speeds in the wind speed group is less than the number threshold, the wind speed group is determined to be an abnormal wind speed group; If the number of wind speeds in the wind speed group is greater than or equal to a number threshold, the wind speed group is determined to be a stable wind speed group.

6. The method for dispatching unmanned aerial vehicles for air-ground collaboration according to claim 1, characterized in that: The calculation method of the response function value of wind speed at each moment is: ； In the formula, For the The response function value of wind speed at a moment, For the Time to The sequence of all wind speed data between the times, For the The wind speed at that moment, The natural constant The exponential function with base , is the mean function, STD is the standard deviation, where the standard deviation reflects the volatility of wind speed.

7. The method for dispatching unmanned aerial vehicles for air-ground collaboration according to claim 1, characterized in that: Establish a mapping relationship between the stable wind speed group and power consumption, including: Obtain the power consumption and normal wind speed at each moment in the stable wind speed group, and obtain a scatter plot of the power consumption and normal wind speed at each moment; The spline fitting tool in the MATLAB tool is used to fit the relationship curve between normal wind speed and power consumption, thereby establishing a mapping relationship between power consumption and normal wind speed.

8. The method for dispatching unmanned aerial vehicles for air-ground collaboration according to claim 4, characterized in that: The K-means clustering method is used for clustering.

9. The method for dispatching unmanned aerial vehicles for air-ground collaboration according to claim 2, characterized in that: The empirical value of the safety threshold is 0.

3.

10. A UAV dispatching system for air-ground collaboration, comprising a processor and a memory, characterized in that: The memory stores a computer program, and the processor executes the computer program to implement the drone scheduling method for air-ground collaboration as described in any one of claims 1-9.

Images